Christine, Marta and Patrick have made a thorough work to try to standardize ACCEL measurement data from before August 2005 and to fill them into the data base. It was found that reference systems are not similar and may have changed with time. It is not possible to find the original measurement files for check, they have been overwritten by ACCEL. The conclusion was that this data should not go into the data base because of difficulties to know what the data really meant (references etc.), and we should continue as before to use the excel files. Flange and service line angles, which are important for studies, cannot be retrieved at all.

For the new template we should add the angles to the upload template and complete the data base from August 2006 with this data.

Patrick states that the confidence level of the old measurements corresponds to the more recent ones.

Sagitta of shells (Gilles Berard):

The sagitta calculated from the two half shells by Gilles is not correlated to the sagitta at itp20. Elena saw some strange mean value shift in the files sent out by Gilles and asked Gilles to recalculate the mean. If the mean has a positive shift this can explain (still conjecture) the sagitta change between industry and CERN. Gilles  re-checked his results and will send the new file.

Saw tooth problems for the production (Zhengwei Gao):

Zhengwei showed that many, in particular Firm 2, magnets have a change in flange position when measured at CERN. The problem may be due to saw tooth in industry or at CERN. The selection of Zhengwei (saw tooth larger than 0.4 mm) will be subject to studies.

Best fit on the magnetic axis for SSS (Jerome Bauquis):

To see the impact of the best fit of the cold bore tube on the movements of flanges and service lines, best fit on the magnetic axes is done. The idea is that the best fit on the magnetic axes is better correlated to the assembly and measurement references in ACCEL. The study was made only on the quadrupole. No significant change in the on beam tube flanges positions could be seen. However, to complete the study it would be needed to look at the magnetic data and check the data points and then add also the corrector magnets (not dipole) to the fit.

'Saw tooth effect' compensation for geometric axis  (Elena Wildner  This is the note from which copies were made as slides):

A conjecture for compensation of the saw tooth effect was shown applied to magnet SSS249. According to Marta, there exist many examples where the method does not apply, those magnets will be tested. The method has been tested for a limited number of cases and the test should be extended, the method refined and better understood. The consequences of the saw tooth has to be evaluated statistically on the machine. Patrick wanted to point out that the bad measurement quality in the example is not representative for SSS measurements and this was agreed on. This was just taken as an example how to analyze and understand a bad measurement and get good information for judging the geometry. To be followed up next meeting starting with a general statistical overview of the dipole saw tooth (Natalia).

Patrick will set up an experiment and measure one point outside the tube with Leica beam going through the tube and with Leica beam not going through the tube. The two measurements will necessarily be put together via bundle.

AOB

Patrick wants to discuss the uploads of ACCEL data to the MAS database (next meeting)

Next meeting: Thusday in a week.

Elena